Editor's Note: For this special issue, we asked our columnists to cover a variety of e-business topics. Their insightful commentary provides a well- rounded outlook as to the benefits and challenges of the e-world. Regular column format will return next month.
From how many sources does the average business collect data? There's real-time online and store data, customer data housed in different sources because of mergers and acquisitions, and data on diverse product offerings manufactured at different plants. All these sources use different databases and operating systems. These are just the tip of the iceberg; the list could go on and on. The extent to which your organization harnesses this data and coalesces it into a uniform feed of information to support a customer-centric business focus will largely determine how successful it is in the e-business marketplace. It's not enough anymore to piece reports together from different organizational systems whether they are multiple operational data sources, independent data marts or warehouses, ERPs or combinations thereof. Just because data from different systems happens to be on one page doesn't mean it has a common source, meaning or accuracy level. To achieve that, you need a strategy not just a toolset of enterprise application integration (EAI).
Enterprise application integration as a strategy means understanding that customer interactions are not isolated transactions that require a linear flow of information. They are events that trigger a series of processes, and those processes need to drive what applications are used and how problems are solved. For example, customer interaction in a partially or linearly integrated organization might go something like this: a customer contacts the call center of an etailer to inquire about the status of an order. The call center representative uses a Web browser to check the status of the order. This front- line user action initiates a front-to-back-office chain of events that is a single activity following a linear pathway that relates only to that particular order and customer. To be sure, all applications such as sales, inventory, order management and distribution, etc., must communicate to conduct the activity; but it still only constitutes one task. It is one transaction. There is no integrated view of the customer and no opportunity to further the customer relationship.
The same transaction, within an organization that uses EAI technologies, may go something like this: the call center person uses the Web browser to view holistic customer information, not just that one order. The current order information is, of course, on the browser screen, but so is other historical and demographic information. The call center representative has access to combined information from many systems to feed the business process of customer service. If there's a problem with the order, the call center person can suggest other product offerings based on past purchasing data or alternative shipping methods if shipping is the bottleneck. Even if there's no problem, empowered employees can use the customer inquiry to perform cross- selling or up-selling. It's a subtle shift; but instead of thinking of a customer transaction as a discrete event, the paradigm is now to think of the phone conversation and order check as an instance of the process of customer service a process that pulls together financial, marketing, sales, customer data, inventory and shipping information to create an opportunity to establish or enhance a customer relationship.
Of course, any strategy implementation requires tools. There are several different tool development philosophies which have engendered products that fall under the rubric of EAI. Some are repackaged middleware tools that have been dressed up with new communications standards, but the best-in-class EAI tools approach the problem differently. Currently, there are three exciting frontiers in EAI tool development: application switch, data transformation and XML authoring standards. These technologies are leading the way in helping companies implement fully mature EAI strategies.
Application switch technology uses intelligent adapter tools to integrate diverse information systems. They work by passing messages back and forth between the native platforms and information brokers. They are "intelligent" because they can be configured to detect any new functionality added to the integrated applications. There are four basic types: database adapters, language adapters, mainframe adapters and application adapters.
Database adapters integrate applications by mapping transactions to a query language usually SQL and thus to the database. Language adapters integrate applications where the programming code is not proprietary in other words, it can be modified by the licensee. Mainframe adapters integrate hardware and software for mainframe systems. Finally, application adapters integrate applications where the code is proprietary where it cannot be accessed or modified by the licensee. If you are using this technology, you will undoubtedly need all four types, and the maintenance load can get heavy. However, application switch technology, because the adapters are intelligent and can detect and adapt to new functionality, is great when scalability is your main concern.
Data transformation technology combines the ability to work with batch and message data in a single product. Software that uses this technology enables organizations to transfer and integrate data between disparate systems with batch data migrations and message sequences. The transformation tool captures application data, which is transferred into a cache attached to the organization's meta data repository. The data is then transformed by the transformation tool using the organization's meta data into the appropriate format for each destination application.
What this means is that you can develop an interface between an inventory-service system written in FOCUS, a billing system written in DB2 and a Web-based customer service system written in Java and transfer data back and forth between those systems seamlessly regardless of formatting issues inherent in each system. The down side of this technology is that it can be likened to a super-database for the corporation, because it sits on top of existing applications and pulls data from them. So, if your enterprise data model is not adequate or your existing systems don't fill your needs, the data transformation tool will probably not be effective. In other words, the underlying legacy systems must be meeting your needs well to use this technology.
The final and perhaps most promising new EAI technology on the market is extensible markup language (XML) applications programming. XML is not a language, per se, as HTML is. It is a low-level language used for constructing other languages, much like Assembler XML or machine level language was used to construct the ubiquitous C programming language. XML is used to manipulate data, not to input or output. The language cannot be used by itself to construct Web applications, but it can be used to construct languages that permit data transfer between disparate applications. It does this by requiring users of XML technology to build schemas for XML- derived languages. The schemas are just like database language schemas in that they specify the format, grammar and syntax of the language. Once the XML- derived language is written, it can be read by other applications that have the same schema or are able to read the schema.
The downside to this technology is that it's new, and many vendors don't fully understand or support it. The upside is enormous, however. It promises seamless data transfer between applications without tool overlays. The applications will actually be able to communicate at the application level without middleware they will only need to be able to read various XML schemas.
The upside to EAI, no matter which technology you choose to obtain it, is also enormous. Thinking of EAI as a business strategy enables you to think of your applications and processes as symbiotic units functioning together to grow the business. EAI tools eliminate much of the confusion inherent in trying to mesh multiple information systems into one cohesive, optimally functioning enterprise information system. They enable you to rapidly retrieve and coalesce data from all your platforms and systems via a single interface if you wish. Finally, an EAI strategy reduces or in some cases eliminates the need to conduct a wholesale replacement of the organization's legacy systems, which was the unrealistic and ultimately impossible goal of many IT projects in the mid to late 1990s. EAI, as a strategy executed with top- notch tools, delivers on the promise of integrated information for increased revenue.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access